Practice effect of volunteers in repeated psychometric testing. How to handle this intervening variable in clinical...

Performance in psychometric tests may show a practice effect with repeated testing. Ideally, a plateau of efficiency should be reached prior to first drug intake. In order to assess the period of familiarization in a multiple choice reaction task (MCRT) 17 healthy subjects practiced the test up to 36 times in the drug free run-in period of 2 trials. A tentative "inclusion criterion" was used in 7 subjects to decide whether their learning phase could be regarded as being finished. Seven to 20 training sessions were required for a first rapid learning phase. There were remarkable differences in the absolute performance of the subjects and in the time course of their practice effect. However, the time course of the learning phase was not dependent on the absolute performance level. By means of the tentative "inclusion criterion" it was possible to reject a subject with a low but highly variable performance. However, it was not stringent enough to indicate the final plateau of efficiency. More "drug free" data concerning the practice effect should be published to enable the users of psychometric tests to make a reasonable selection according to their needs and to help them design their trials properly.